Authors: Fujin Zhou and Remco Oostendorp
Vrije Universiteit Amsterdam
March 2013


* Outline: this file provides instructions on which programs to run in which sequence

Step 1: run Stata dofile "1exportdataforuseinOX.do" to export the following data
        from dataMongolia.dta: 0IC.csv, 1ICa.csv, 1ICb.csv, and 2IC.csv
Step 2: run the 4 OX files 
  - MIMICbyMLE0no-d1.ox, using data 0IC.csv
  - MIMICbyMLE1no-d1A.ox, using data 1ICa.csv
  - MIMICbyMLE1no-d1B.ox, using data 1ICb.csv
  - MIMICbyMLE2no-d1.ox, using data 2IC.csv
  and produce Table 3 in the paper
 
Step 3: run the Stata dofile 2importOXoutput.do, 
        which imports the prediction results by MIMICbyMLE2no-d1.ox 
        and save as output2.dta,
        and then merge it with dataMongolia.dta, save as 3underreport.dta

Step 4: run the Stata dofile 3compare3approaches.do, produce tables 1, 2, 4 and 5 in the paper
        
appendix:

input data
======================
1. dataMongolia.dta

output data
======================
1. 0IC.csv
2. 1ICa.csv
3. 1ICb.csv
4. 2IC.csv
5. output2.dta
6. output2.csv
7. 3underreport.dta

Log files
========================
1. logimportOXoutput
2. logcompare3approaches
3. OxOutputforMLE0no-d1
4. OxOutputforMLE1no-d1A
5. OxOutputforMLE1no-d1B
6. OxOutputforMLE2no-d1
